Review of low voltage load forecasting: Methods, applications, and recommendations

Haben, Stephen; Arora, Siddharth; Giasemidis, Georgios; Voß, Marcus; Greetham, Danica Vukadinović

doi:10.1016/j.apenergy.2021.117798

“…Several residential electricity consumption datasets can be found in the literature, each of which with its own characteristics as summarized in different survey papers 12 – 14 . Such characteristics include the number of sensors (e.g., a single sensor for the whole building, circuit-level, and appliance level), type of measurements available (e.g., current, voltage, active and reactive power, and energy tariffs), data granularity (e.g., from several kHz to one sample every hour or less), and dataset duration (e.g., from a couple of days to several years) 15 .…”

Section: Background and Summarymentioning

confidence: 99%

A residential labeled dataset for smart meter data analytics

Pereira

¹

,

Costa

²

,

Ribeiro

³

2022

View full text Add to dashboard Cite

Smart meter data is a cornerstone for the realization of next-generation electrical power grids by enabling the creation of novel energy data-based services like providing recommendations on how to save energy or predictive maintenance of electric appliances. Most of these services are developed on top of advanced machine-learning algorithms, which rely heavily on datasets for training, testing, and validation purposes. A limitation of most existing datasets, however, is the scarcity of labels. The SustDataED2 dataset described in this paper contains 96 days of aggregated and individual appliance consumption from one household in Portugal. The current and voltage waveforms were sampled at 12.8 kHz, and the individual consumption of 18 appliances was sampled at 0.5 Hz. The dataset also contains the timestamps of the ON-OFF transitions of the monitored appliances for the entire deployment duration, providing the necessary ground truth for the evaluation of machine learning problems, particularly Non-Intrusive Load Monitoring. The data is accessible in easy-to-use audio and comma-separated formats.

show abstract

“…Short-term forecasts can further inform participants of local and peer-to-peer energy markets [7], real-time pricing schemes [8] or flexibility applications [9] that are emerging with the energy transition. See [10] for a review on more applications.…”

Section: A Motivationmentioning

confidence: 99%

“…Whereas usual point forecasts model the expected value, probabilistic forecasts model the distribution, enabling more informed decision-making by considering uncertainties. This makes probabilistic forecasts especially relevant for the LV level, as they can better handle the higher volatility present at this level when compared to the systemlevel [10], [12].…”

Section: A Motivationmentioning

confidence: 99%

“…Generally, uncertainty in load forecasts can be modeled as either simple prediction intervals [14], quantile estimates [15]- [17], full continuous distributions [18], [19] or scenarios [20], see [10] for a review. Further, the applied methods can be distinguished between statistical and time series models that often rely on parametric assumptions and non-parametric models from the machine learning domain.…”

Section: B Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Short-Term Density Forecasting of Low-Voltage Load using Bernstein-Polynomial Normalizing Flows

Arpogaus¹,

Voß²,

Sick³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

The transition to a fully renewable energy grid requires better forecasting of demand at the low-voltage level to increase efficiency and ensure reliable control. However, high fluctuations and increasing electrification cause huge forecast variability, not reflected in traditional point estimates. Probabilistic load forecasts take future uncertainties into account and thus allow more informed decision-making for the planning and operation of low-carbon energy systems. We propose an approach for flexible conditional density forecasting of short-term load based on Bernstein polynomial normalizing flows, where a neural network controls the parameters of the flow. In an empirical study with 363 smart meter customers, our density predictions compare favorably against Gaussian and Gaussian mixture densities. Also, they outperform a non-parametric approach based on the pinball loss for 24h-ahead load forecasting for two different neural network architectures.

show abstract

“…Several residential electricity consumption datasets can be found in the literature, each of which with its own characteristics as summarized in different survey papers [12][13][14] . Such characteristics include the number of sensors (e.g., a single sensor for the whole building, circuit-level, and appliance level), type of measurements available (e.g., current, voltage, active and reactive power, and energy tariffs), data granularity (e.g., from several kHz to one sample every hour or less), and dataset duration (e.g., from a couple of days to several years) 15 .…”

Section: Background and Summarymentioning

confidence: 99%

A Residential Labeled Dataset for Smart Meter Data Analytics

Pereira¹,

Costa²,

Ribeiro³

2022

Preprint

0

View full text Add to dashboard Cite

Smart meter data is a cornerstone for the realization of next-generation electrical power grids by enabling the creation of novel energy data-based services like providing recommendations on how to save energy or predictive maintenance of electric appliances. Most of these services are developed on top of advanced machine-learning algorithms, which rely heavily on datasets for training, testing, and validation purposes. A limitation of most existing datasets, however, is the scarcity of labels. The SustDataED2 dataset described in this paper contains 96 days of aggregated and individual appliance consumption from one household in Portugal. The current and voltage waveforms were sampled at 12.8 kHz, and the individual consumption of 18 appliances was sampled at 0.5 Hz. The dataset also contains the timestamps of the ON-OFF transitions of the monitored appliances for the entire deployment duration, providing the necessary ground truth for the evaluation of machine learning problems, particularly Non-Intrusive Load Monitoring. The data is accessible in easy-to-use audio and comma-separated formats.

show abstract

Review of low voltage load forecasting: Methods, applications, and recommendations

Cited by 79 publications

References 173 publications

A residential labeled dataset for smart meter data analytics

A residential labeled dataset for smart meter data analytics

Short-Term Density Forecasting of Low-Voltage Load using Bernstein-Polynomial Normalizing Flows

A Residential Labeled Dataset for Smart Meter Data Analytics

Contact Info

Product

Resources

About